Procrastination, Obedince, and Public Policy: The Irrelevance of Salience

lly GARY M. ANI>IXSON arid WAI.’I‘I:R ~ o C K *

AIHKACT. Modern economics assumes that individuals are rational maximizers, who, i n the presence of costly irzjounzutiorz, sometimes make mistakes. Recently, George Akcrlof has challenged this presumption. He proposes that people sometimes fail to maximize their long-run utility, due to the phenomenon of “salience.” This refers to the alleged systematic distortion in individual per- ception where events closer to a person (in time and in space) seem bigger and more import:iiit than they really ;tw. ’I’he sulicrzcc plienomcnon is claimed to interfere with the Ixocess of rational mu3iitizizufiotr, rendering some individual choices, in effect, irrational. This p q m subjects Akerlof‘s suggestion to critical scrutiny. I t is argued that the exainples Akerlof oflcrs o f the elfect of salience :irc not thr :tnoinnlies Iic diims lxil i i i s t d rrprcsrnt bc1i;iviors which c:tn IIC rcaclily explaincd within the I‘ramework of the stanclarcl economic model. All 0 1 I I / S I I I ; I ] O ~ cxmplcs itre coiisltlc.rccl i t i ibis llgltt :itid It Is sirggcstecl that the work of Isrucl h’irzricr holtls more promise for improving the :tii:tlytical power o f the st:intl;trd economic model th;m does I he idea o f salience.

1

Introduction

Tiit I D E A that individuals fail to evaluate ;iccuratcly the real value o f future pros- pects has a long history.’ Many scholars have felt that individuals systematically fr t i l to evaluate correctly the present value of fu ture prospects.

Recently, this venerable idea has received a refurbishing from a distinguished modern economist, George Akerlof. He argues that economists should pay greater attention to a central principle of modern cognitive psychology, “sa- liency.”

This term refers to the purported tenclcncy o f individuals to place greater cognitive weight on events which strike them as especially “vivid.” But a more “salient” event may not necessarily be iiiore important or significant in the individual’s longer view; it is merely more striking, or memorable for reasons essentially unrelatecl to its “true” signific:tnce.

[Gary M. Anderson, PhD., is professor of economics at California State University, Northridge, CA 9l.330 :tnd W;ilter Hlock is professor of economics a t Iloly Cross Collc~e. Worcester, MA O l G l O . ]

Anirrirnn Journal of liconimics and Sociology, Vol 54. No 2 (Apri l . 1Y95) 0 1W5 Ameri<dn Jc,iimil of Economics and Scn‘iology. Inr

202 Americaii Journal of Economics and Sociology

We contend that the basic thesis offered by Akerlof is unsound, and tlie many public policy recommendations he proffers cannot be justified on its basis. Our purpose in the present paper is to examine critically the proposed role of “sa- liency” in economic affairs. We will argue that this bold attempt to reconstruct modern economics is i n reality a disappointing tizzle.

The paper is divided into eight sections. Section I 1 outlines the thesis of Akerlof‘s article :mtl exp1:iiiis his posiiion on the issue o f “s;ilicncc” i n human affairs. Section 111 discu s some general, Iusic criticisms of Akerlof‘s itlea. Section IV ex;iniincs a m a j o r Akerlof applic;ition, his ;in;ilysis o f drug :iclclic- tion. Section V considers Akcrlof on another iniport;int application, olxdience to authority. Section V1 discusses Akcrlof‘s cxtcnsion of his s:ilicncc principle to the problem of cults. Section VI1 acldrcsscs tlic implications of his argument for public policy generally. Finally, Scmion V I I l concludes o u r argument.

11

Akerlof on Saliciice arid 1’rocrantlll;ltloll

Tile A K I ~ R I . O I : I I I ~ ~ ~ I S is I)olcl, yet simple i n f o r m . I t holds th;t t i n ccrt;iin situ;ttions involving a long scrics of snxill incrcmcni:il rcp,;iicd tlccisions, p c ~ ) p l c ~ ;ire likely, all too often, t o p1;ice undue emphasis on the iniport;ince of present events, and to cliscount fuiurc ones too Iicavily. “ ~ ’ ~ C S C I I I I)ciicliis : i n d c o s t s may have undue salience relative to future costs a n d Ixwefits” (Akcrlof 1991, 1 ) . Given this “dynamic inconsistency” ( 4 ) , people will not maximize their “ ’true’ utility” (2 ) . He argues that this discrepancy between perception and reality may help to explain sever;il seeniingly tliffcult p1icnorncn;i conin~on in cvc~-~day life. For example, “procrastination” may occur when present costs are “unduly salient” in comparison with future costs, “leading individuals to postpone tasks until tomorrow without foreseeing that when tomorrow conies, the required action will be delayed yet again” ( 1 ) . Similarly, “irrational ohedience to au- thority” may occur when the salience o f an action tod;iy clcpcncls upon its de- viation from previous actions.

Akerlof believes that his theory of s;iliency hiis practical :inti important im- plications for public policy. Sometimes pcoplc just clon’t know what is good for them, in their own ternis as defined by their personal utility functions:

Individuals wliosc hchavior reveals the v;irious patliologic~s I sIi;ill niodcl :ire iiot m:ixirnizing their ‘true’ utility. The principles of r i l cd prcfcrcncc c:innot tlicrcforc be iisccl 10 assert

that the options that are ~hoseii must Ix preferred t o thc opiions t l i : ~ ~ :ire no^ choscn. Intlivicluals may be made h e w r olT if thcir options :ire limited :ind their choices coiis~r:iinccl ( 2 ) .

He offers several examples of such “utility enhancing” coercions. 1:orccd pension plans may be superior to voluntary pension schemes; prohibitions on alcohol

Salience 203

and drug use may make potential users better off than merely taxing intoxicants; and a key role of management may really involve the (arbitrary) setting of schedules and deadlines, which presumably help their employees to discover their “true” productivity. Thus, according to Akerlof , coercive paternalism can often improve the well-being o f the coerced parties.

Describing this theory as “bold” is an understatement. Taken literally, it rep- rrsents ;I r:iclic:il rejection of the neoc1:issical economic moclcl of m a n ns a r:i-

tional, sell-interested ui i l i ty maximizer. ’I’he standard economic model ac- knowledges 11i;ii int1ivitlri;il c.1~oicc.s m i y sonwiimes be tnist;ikcn, when f o r ex- ample :I person makes :I selection I~asc~cl on incomplete o r faulty information. The intliviclual is still ni;iking :I consistent d T o r t t o ni;ixiniizc his utility even though his :ictual pcrforni:incc Ic;ids to t1is:ip~)ointing results. Instead, according to Akerlof, oftentimes inclivicluals just prefer the “wrong” things in terms o f their own welfare. Sometimes individuals f a i l to niaximize their utility i n the long riin.

In devoting careful attention to the role o f “salience” in human affairs, “ p r o - cr:istinai i o i i ” is, l)c*rli;ips, ilic 1)riiii;iry i l lus i r ; i t io t i Aker lo f ofi’crs. “I’rocr:istin:iiioii’’ rcfcrs to the phenomenon where ; in intlivic1u;il delays taking action with respect t o some matter in spite of an ;iwareness that a prompt response would I x superior.

I t is cotninuii I’or clelaycrs to proc1:iitn to otlicrs tliat ilicy “me;int to” do this or i1i;ii sooncr, implying t1i;it they did not i n f:ici m:iximizc. I n Akcrlof’s view, i n effect, individual discount rates are too high i n cases where a pattern of regular decisions milst be niaclc, and time is short between each discrete choice; where a penalty, o r a “salience cost” is imposed by making the wrong decision; and where ration:il expectations do not apply, in that the actor “did not foresee t1i;it when tlie next day came (lie) would continue t o

put off the decision for an additional day” ( 3 , 4 ) . Thus, an individual in such a circumstance exhibits a failure to r:itionally maximize utility across time. ?‘lie vividness of the experience of the moment dominates, and the individ- ual’s long term interest-defined i n that individual’s own terms-suffers as :i result.

Akcrlof tlcscrilxs such behavior as irrational, Le., meaning that i t represents a siinplc failure by the person t o maximize his own utility. I t is not merely that the person possesses inadequate information regarding the future consequences of tl;iwdling, but rather that he simply fails t o compare correctly the costs versus the Ixnelits o f procr;isiinating. ’l‘hus, tlie Ixhavior Akerlof means to describe cannot merely be dismissccl on information cost o r transaction cost grounds. I’eoplc ;ire not merely prone io niistakc~s I~ccause they h;ivc ;iccess only to irn- perfect informition, but ;iIso Ixuuse they sometinics process the available in- formation i n the wrong way.

i

204 American Joidriial of I:co?iomics arid Sociology Salieuce 205

111

Was Akerlofs Analysis of Procrastination Worth Waiting For?

IF AKERLOF IS CORRECT in asserting that this kind of behavior is common, this fact has radical, and troubling implications for economics. In effect, if people often behave irrationally then much human action is simply unpredictable and beyond analysis.

Therefore, a simple question has iniport:int inip1ic:itions for economics: do we have reason to take the (aclniittcdly) coninion representations made by individuals concerning their own “irrationally I~rocr;istinating” behavior at face value?

I t is curious that Akerlof, in the course o f detailing his theory o f how rational maximization fails i n the face of saliency, implicitly relies on I I ~ C perfect accuracy of individual self-representations. Simply stated, Akerlof seems to assume that people always tell the complete t r u t h when descrilJirig their niotiv;itions.

But real people often hi1 to lcll t1ic complctc t i - ~ i t l i : i l x ) \ i t Ilic‘ir Iwrsowl nio -

tivations as well as everything else. Beyond the obvious fact that people some- times lie (intentionally misrepresent the facts as they unc1crst;ind them to be), much human time aiicl :irtcntion involves :itteiiipts to Ixirgaiii with others. ‘I‘his bargaining extends to all human interaction. I n many circunistaiices, complete and total honesty will not be the optinial strategy i n a 1):irgaiiring situation.

This consideration applies to an iniport;int il1ustr;ition u d I>y Akerlof, the supposed “failure” of a diet to lead to actual weight loss.

Dieting represeqts a prc-ciniinently social activity. N o t o n l y is an improved personal appearance i n ihc eyes of othcrs typic;illy tlic obtciisil)le go;il o f the diet, but the diet-related behavior can also be a kind of performance designed to serve some other purpose than weight loss. The very fact that others are aware of a person’s diet suggests that the person has undertaken some investment designed to communicate the fact to those people. This behavior-the overt manifestations of the diet-may, in part, represent a strategy clcsignctl to influ- ence others in the immediate environment; actual weight loss may be only a secondary goal, o r i n actuality not really :I go:il : i t a l l .

Consider a stylized example. A wife expresses dismay to her husband con- cerning his overweight, unattractive appearance. The liusl~ancl goes on a diet in order to appease her, perh;ips only tcmpor:irily. A n extcrn:il ol>scrver of this interaction will likely be unable to ascertain [lie liusliantl’s t rue , as opposed to the stated, goal. i t is possible that rccluccrl tlomcstic strife, and not weight loss, is the true goal of the diet. Thus, the failure of the diet :IS a weight loss tactic may simply signal that the true objective was clifl’crent than the goal the dieter

represented to others. Dieters, like the rest of us, sometimes find i t expedient to misrepresent their true goals in pursuit of strategic benefit.

Misrepresentation may not necessarily include actual lying. Losing weight may indeed have been one of the dieter’s goals. But it might not have been the only goal, or even the major goal.

This principle of systematic misrepresentation applies t o the phenomenon, procrastination, more generally. Procrastinators are often individuals who dis- agree with other people about the correct prioritization of behaviors, but who prefer t o portr:iy this clisagrcemc~nt :IS sonic*tliing else. “ I nic;int to” ancl “ 1

couldn’t help myself” sometimes stands in for “ I did not want to, and I refuse to x h i t the kict.” I>islioncsty may not lie ciliical, Iwt i t can 130 eminently rational.

Alternatively, ;I Ixliavior which might sceni i r r a t i o n a l o n the surface, might merely be evidence of shifting patterns i n preference ordering. Consider once again the example of the apparently [ailed diet. A person might genuinely want to lose weight, and at the same time want to e;it delicious but fattening food; sometimes, seemingly, lie wants both at the same time. ’I‘lie person must choose I)etwceti tlic ~:itisl:ictioii I i o n i iii:iiiii:iiiiiiig ;I sllrii 1)otly :u id the pleawres ( 1 1 eating fatty foods. A rational indiviclual will necessarily rank one or the other of these conflicting goals higher than the other at any given time.

’i’lie st;iiid;rrcl ccoiioiiiic iiioclcl ;issunies t l u t inclividual preferences are coil- sistent 31 any inonirnt, yct not ncccss:irily const:int o v r r timr. The prioritization inclividuals place on conflicting go;ils may shift over time. Thus this year I prefer slimness over the enjoyment from eating fatty food, so m y diet works and I lose weight, but last yc:ir gastronomic tlcliglits took prececlcncc over my desire t o

avoicl obesity, and I gained weight. ‘I‘his shifting o f preference ordering, com- bined with strategic consiclerations leg., m y desire to minimize marital conllicts causes me to claim that 1 have been on a diet all along] seenis sufficient expla- nation for my observed behavior without the need to introduce any new prin- ciples like “salience.”

Thus, in our view Akcrlof‘s attempt to explain procrastination by way of his theory o f saliency is unnecessary and overwrouglit. ’I‘hc Ixhavior he offers as supposeclly 1mxloxic;il, is i n reality easy to explain within the confines of existing economic tlieory.

One of the most basic insights of modern economics is the subjectivity of preferences. I’riccs and c1u;intitics clcmanclccl x i c l supplied ;ire objective facts, nie:isur:il~le mtl, i n some sense, rcli:il~le. On l l ie oilier Ii:ind, the true preference orderings of individuals are subjective and cannot be directly measured by out- side ol)scrvcrs. Coriscclucntly, the assert ions o f (self- interested) partics regartling their personal preferences ;ind motivations arc inherently probleniatic. One ignores this simple methodological precept at one’s analytical peril.

206 American Jouriral of Econoniics and Sociology Salience 207

However, Akerlof is not merely intent on proposing a radical new principle in the analysis of human behavior, but also intends that this principle has quite policy-relevant implications. These claims will now be examined in detail.

IV

Addiction and Procrastination

AKERLOF APPLIES 111s MODEL of procrastination to drug abuse, and explicitly rejects the Chicago School's ration;ilistic model of ;idtlictivc 1wIi:ivior (Ik*ckcr ; ind M u r - phy 1988; Stiglcr and Uecker 1977). "Most drug abusers, like most chronically overweight intliviclu:tls, fully intend to cut clown tlicir ini:ikc, sinccs they recognize that the long-run cost o f tlicir ;tdcliction cxccc~ls its bcnclits," Akcrlof writes (1991, 5). But the "certain and immecliate" (salient) rcwarcls win out over the possibly lethal (nonsalient) costs.

I t is important to recognize what Akerlof is not arguing. I l c 1~:ises his defense of government anti-drug policy on a purported failure of tlic incliviclual drug consuiiier to r:itioii:illy iitmiiiii7.c. his ow11 ui i l i iy , ; i i d tlcrcs iioi iit;ilic :illy c,l:iirtis concerning externalities purportedly gencratcrl by the consumption o f drugs. Drug use may or may not produce dctriment:il impacts on society. Ilut Akerlof takes no position on this question.

We readily grnnt that indivit1u;ils with scriows, physic:iI, xklictions to drugs do not seem to be plausible candidates for the role of ration:il ut i l i ty maximizer. Some addicts do indeed seem to engage i n fatal activities f o r what seems to an outside observer paltry short-term satisfnction, :ind :I r:ition:ilistic cxp1an:ition for such extreme cases is hard to accept. Even i f we assunie, however, for the sake of argument that such Ixhvior is actually irratioixil, just how big and ex- tensive is the problem?

The answer seems to be, fairlysmall. The long suffering addict is a comparative rarity. Most drug users seem to indulge occasionally, as a form of recreation. Most users of intoxicants (including substances like alcohol) seem to maintain homes, jobs, and families with apparent success. Akcrlof exaggerates when lie suggests that most "substance abusers" are comparatively helpless addicts. To the contrary, most drug users seem rather ordinary in all other respects.

One need not accept that all drug users are carefully calculating rational actors to recognize that most drug users seem quite r:ition;il in their conduct. Drug users, like practitioners of many other activities, accept certain levels of risk in exchange for expected benefits. Most drug use involves risk to the user, but does not represent assured self-destruction; moreover, the act of intoxication per se produces pleasure for the user that counterb:il;inces thc cost associated with the expected level of risk. That objective risk might be unclcr or over-

estimated by the drug user, but the model of rational maximization will still apply, subject to information constraints.

Akerlof's willingness to dismiss the drug addict's choices as irrational is un- necessary, hut this does not necessarily mean that drug use can have drastic implications on the individual's utility maximization problem. Some drug addicts are, i n fact, different after their experience of using drugs.

l'he experience of intoxication itself can, indeed, lead to a kind of er ror on the part of some drug users. The problem relates to the fact that intoxication is inhcrcntly rlillicult t o ;inticip;itc i n clcl:iil for int1iviclu;ils un:tccustoined t o i t . Not on ly is i t possible f o r some drug users to unclerestimate the pleasure associated with tlic cxpericncc, 1x11 the clrug expcricncc seems t o ;tllect the unclcdying utility I'uiictions of at least some users. 'I'hc persoii who lakes drugs may exhibit significant changes in 1xxson;il priorities after, as compared to before, the ex- perience. Sometimes, the shift may be sudden and the resulting dislocation and tliscoort1iii:ition o f i h : i i person's plms might be extrcnic, :it least temporarily. This possibility reflects ;I kind of error i n decision-making but does not imply I t - ~ i t ii)ii:i I i i y .

Akerlof misses a n important problem with the Chicago theory of rational ;iclcliction concerning the role o f error in clecision-making. Kirzner (1973) points to a dilliculty i n the Chicagocsque world o f Stiglcr (1961) and 'I'elser (1966), whcrc no on(' cmi evcr niisc:ilctil:itr. In ihc Stiglc-r.'l'clscr iiiform;ition-sc;ircli model, genuine error is impossible; i f it is later shown that an expost mistake w ~ s made, the defence can always be tliat ex utile the information about the good, :ind ilir ~ o o d ilsc-lf wcrc, i i i clTcct, :I joint product, :tncl 11i;ii tlic c~xpectetl utility of another day's search was less than the expected benefit. Telser (1966, 458) argues that selling a car which inclucles a carburetor is like selling a car together with information about the car. Kirzner (1973, 156) takes issue with this comparison, and argues tha t the very conception of clemand presumes some knowledge, however limited, of the relevant product. Genuine error in the Kirznerian world arises out of total ignorance of the availability of automobiles, o r other products, not just of their costs and bcnelits.

Ilnderstancling the shortcoming i n these information search models clcpends crucially o n the fact that while demand curves can indeed be drawn for com- modities of whose existence one is aware, it is meaningless to draw a demand curve f o r a procluct about which one is completely unaware. Kirzner (1973, 158) considers the possibility o f ;I person who h:is n o inkling that a particular kind o f commodity exists:

We ni:iy, of coiirw, im;igiiic his (Icni:ind curve for I l l is ronimotlity oiice its exisleiire Ii:is

I,eromc known to Iiini. lhit if we wish t o discuss the roniniodity iii its unknown stare, we ;ire siiiiply un:il,le t o u lk 0 1 the roiisiitiicr's clcm;iiid for i t . I t is not that his demand curve

208 American Journal of Economics and Sociology Salience 209

coincides with tlie price axis; that he would huy none of i t at anygivcn price. 11 is r:cther that the very notion of demand has no place under these circunistanccs. . . With no opportunities perceived. the notion of consumer demand has no meaning.

In other words, the Stigler-Telser information model is quite adequate to the task of accounting for length of search for ordinary, known, commodities. Here, in their sense, no errors are possible. B u t with regard to completely unknown products (e.g., the frislxc o r hula hoop or pet rock or automobile before their respcctive introcluctioiis) the motlcl c;iniioi eveii he :ippliecl, because i t is illegitimate t o utilize c1em:incl curves, : ind wi thout them, the niotlel c:ititiot I,c- ;ipplitd. I n tlic~.sc c;is<~.s, t h v n , t ru( ’ cruir c;in iiitlccd, occ 11 r .

13ut while the (;liicago appro:icli ~iticIerctii~~Ii;isi~.cs the role 01 cwor in eco- nomic :iIT:iirs, Akc.rlof seem to :issumv :I world in w1iic.h tirist;rkc*s :ire tlic rule. Neither extreme is warranted; we need to develop ;iii alteinative theory of ex- pectations, where iiiisc~ilc~il~itioti i h tic4icr virt i i ; i l ly o 1 1 1 1 1 1 1 ) I c ~ c ~ i i t nor totally absent. This more moder;ite position irnplics : in ccoiiomic iitiivcwv whcrc ~~cople can change their minds, continu:illy, :IS i n the C~;ISC of clicis and :iclclictions, and once and for all, or at least gr;idually, as in tlie case of saving money for

In fact, another example of “salience driven” I~rocr:istiiiatioii olrcretl by Akerlof involves interteiiipor;il savings m c I coiisuiiiption clccisiotlh. ’l’lics presciii costs of deferring future consumption ;ire more “s;ilicnt,” and 1c:itl iiicliviclirals to s:ive less than the optimal amount. Akerlof notes t h a t Irving Fibher licld a similar view, and spoke of the “preference for present gratilicatioli” 1)cing powerful because “anticip;ition o f the future is roeak” (qiiotccl i n Akcrlol 1991, 6; emphasis as in the original). Later he states that the hypothesis “ i n the ;il>sence of pension plans, many individuals lack suflicient self-discipline to Ixgin saving for retire- ment in a timely fashion is consistent with the finding that there were high rates of elderly poverty prior to the rapid, unexpected growth i n Social Security pay- ments in the late 1960s and the 1970s” (7).’

This is a prime example o f the irrelevance of “s:ilicncc,” : i t Icxt o f Akcrlof‘s version. IIis application o f his idea t o the pro1)lcrii ol‘ h.iviiig ;iclcls exactly nothing to our understanding of the phenomenon of the discount rate. Other things equal, now is better t1i;in later to the r ; i t i o n ; i l :ictor. ’i’o :isscrt t h a t the present has greater “saliency” compared to the future atlds nothing of sub- stance to o u r understanding of the discount rate. l‘he motlcl o f rational self- interested ut i l i ty maximization is perfectly capa1,lc o f ;ic.cominotl;iting the phenomenon to which he here refers. “Salience” S C ~ I I I S t o I x a term in search of a theory.

old age.

V

Obedience and Salience

TIE FAMOUS EXPERIMENTS of Stanley Milgram receive attention from Akerlof as purported evidence that pathological, irrational obedience to authority is a com- mon behavior. Since these experiments arc cited by Akerlof as supporting ev- idence f o r his contention that saliency leads t o irrational outcomes, the subject deserves a closer look.

Milgrmii recruited a iiuinlw’ of prticip;ints t o his experiments by a ni:iil cir- cul:ir, th; i t f;ilscly listed I I I C pirposc o f the cxpcrimcnt ;IS ;I study of the clrects 01’ I)iiiiixIcIIicIi( on riicbniory. ‘l’lic rc*;il i i i lcvi t W ; I ~ to itivcstig:itc I I I C willitrgtic*sh 01’ rliosc p;trtic,ip;itits i o lollow instrtic.tiotis. I’he c~xperitnctital sihjccts wcrc thcn instructed to ;ict as teachers, :ind ;idminister electric shocks (;IS punishment f o r wrong answers) to ;I conledcrate of Milgram posing as a learner. Milgram fo i i i i t l 11i:it the m;ijority o f llis subjects wcrc’ willing to ;idministcr thc- maximum shock level (in reslity, the switches die1 not really administer any electric shock) to Icbarticrs, cveii tliougIi tlrc sitl~jcct/te;i~~liers were led to 1)clicvc tIi:it such ;I

level of shock was quite painful to tlie recipient. Akerlof interprets this result ;IS supporting his argument that people often exhibit “irrational obedience to :iuihority” (Akerlof 1991, 9).

Akcs r lo f c1;iims 1Ii:it “untluc olx*tlivncc to ; iuthori ty m:iy occur :IS a f o r m of procrastination i f disobedience of an authority is salient and distasteful” (8). Apparently, individuals like the unwitting sulljects in Milgram’s experiment find tlis;igrecmcnt with the cx~~eritiicntcr/;~utliority ligure so dist:istcful that they pnsh the shock lever even though they “know they shouldn’t.”

Akerlof remarks that the Milgrain experiment “demonstrates that isolated in- divicluals can exhibit remarkably obedient (and deviant) behavior inside the lalmratory” (lo).

Nowhere in his article clocs Akerlof make reference t o any difference between coerced and voluniary o1)edience. Surely it matters whether or not an instruction is Ixickecl by coercion o r the t1irc:it of coercion. ‘I’his is a very important dis- tinction, and one which Milgram c1e;irly unclcrstands. Ilc is very clear that his expcrimctit:il subjects were ;dl volunteers, paid for their services (Mi1grdni 1974, 14-16). Although Milgram misrepresented his intent behind employing these suI,jccts, their actua~ pirticipxtioii W:IS compIeteIy voIutit;iry o n their ‘part.

In other words, Milgram (or his employee who conducted the experiment) was not an authority who enforced his clccisions with brute force. Milgram’s expcrimctitcr merely succceclccl in pc*rsu;iding (some) subjects t o follow a rec- ommended course o f action. Itational individual maximizers will often choose to rely on information provided I,y others possessing speci;il expertise-authority

210 Americari Jourrial of Ecotioniics a?zd Sociology Salience 211

figures in the sense that they seem worthy of respect and trust-as a means of minimizing transactions costs.

Consider t h e context of tlie Milgram experiments. The subject recognizes that the entire (purported) test is voluntary for all concerned. 'The student (in reality, one of the experimenters) supposedly subject t o electric shocks meted out following mistaken answers to questions, and who (per Milgram's instruc- tions) loudly complained without ever getting up and leaving, has voluntarily entered the test situation. This person was free to leave at m y time, as tlie real test subject was well aware.

Milgram's main linding should have surprised no one. I le cliscoverccl that complete strangers feel little conipunction ;iI>out annoying each other, :it least in a setting where :il l aspects o f their interaction is voluntary ( i e . , ciilicr party is free to just get u p and leave at any time). Whatever the v:ilue of Milgram's experiments, they are simply irrelevant to the use to wliicli Akerlof puts them.

VI

I'he Cult of Sallence

AKERI.OI: DEVO'I'ES considerable attention to tlie purported :ipplic:il>ility of his theory of saliency to tlie pro1)lcm o f cults. 'l'licsc religious organizLitions are portrayed as prime cxamplcs of t l i c f;tilurc of indivitlu;ils r;i[ion:illy to ni:iximizc when they refuse to qiicsiion authority. 'I'liis matter, too, deserves close attention.

While to the dictionary, a "cult" is a "system of religious 1)clief and ritual," this neutral definition is misleading. Ever since tlie Jonestown mass suicide in 1978, cult has become a term of opprobrium, connoting mental instability and bizarre practices. Akerlof applies his theory of s;ilicncc to cults in this sense since it is clear from his examples that lie means pathological group behavior. Thus, the very act of joining a cult is implicitly presented as a fundamentally irrational act, meaning that to do so tends not to lie i n the individual's long- term interests. Rational, utility maximizing individuals will, therefore, tend to avoid cults.

Akerlof implies t h x tlie act of joining a cult with rigid internal rules constitutes prima facie evidence o f shortsighted irrationxlity on the part of that individual member. But as Iannaccone (1992) denionstrxes, there is a simple and reason- able explanation for such behavior. Cults provide services to their inemhers, such as intense fellowship as well xi more tangilAt collcwive iinclertakings, which are potentially subject t o free riding by members possessing low levels of commitment to tlie enterprise. Strict intern;il rules scrvc to mitig;ite this free riding problem, and improve tlic ability of such org;iniz;itions t o supply collective goods to their meml>ersliip. Cults that control frcc riding inorc efficiently are

more likely to survive in competition with other groups who offer less stringent internal rules. Thus, the popularity of relatively strict cults has a rational expla- nation.

The process of joining some organizations, religious and otherwise, often involves a series o f stages, and Akerlof regards this gradual acceptance as an example of irrational acceptance of authority. Consider the Unification Church, one of Akerlof's principal examples. He outlines tlie elaborate process of in- duction into Unification Church confronting new adherents:

Coiivcris mikc :I sequence o f sn1:111 dccisioiis to :icccpt :iutlioriry. . . as :I wsuh o f this sequence of decisions to obey r:lIiicr dwii lo rcIxI, the convcrls tlcvclop heliefs and values very dilkrcnl from wliai dicy had ;it tlic Iwginniiig. . . h e clissciii neccssxy for resist;ince t o escalation of coiliinitiiiriit docs no^ develop ( I I ) .

I n other words, Akerlof is concerned that groups of like-minded individuals develop as a result of a voluntary proccss o f self-selection. As lannaccone ex- plains, such behavior on the part o f cults may represent a rational response to ;I free riding problem. I:urtlierniore, all religious orpnizations attempt to increase :iItciid;iiicc :i i i iI nic*tiil)crsliil), i t i orcIc*r to S;IVC souls o r t o do otlicr good works. Many of them do this in the form o f slow easy stages so disturbing to Akerlof. 'This also applies i n tlie commercial world, especially f o r big ticket items, for example by tlie use o f c;isy credit ternis. liven courtship follows this gradual progrmi; tlic m;in huys ~ h c woiiiaii flowers, lie opens tlie door for her, he is polite to lier Iyarcnts, he tries to insinuate himself into lier good graces using a myriad of step-by-step techniques, and so on. In each of these cases, the observed behavior seems to simply be an act of persuasion. As such, it represents a form of rational behavior and requires no new economic principle to understand.

Akcrlof implicitly :rcccpts ;I politicized nioclel o f a non-political form of o r - ganization. Clubs of all kinds normally restrict membership to individuals with certain interests and inclinations, and implicitly (or explicitly) exclude others. I3ut tliose excluded froin a club are not dissenters i n any political sense. 'I'he term, "dissent" implies a pre-existing process of allocating resources in a coercive manner. A stampcollecting cluh may implicitly exclude individuals uninterested in philately, but the latter have nothing t o dissent about. I n a free society, clis- satisfied cult meinlxrs can promptly make themselves ex-members, and real- locate their energies, money, antl time elsewhere-perhaps to the cult down the street! Uy referring to conversion, antl suggesting that tlie absence of dissent represents the outcome o f ; in irr;ilion;il process, Akerlof implies t1i:rt a range o f perfectly voluntary incliviclual activities are really (somehow) politically coerced. Akcrlof's misuse o f these 1o:iclctl terms merely invites unnecessary confusion.

'I'lirougliout his discussion of tlic Ilnilication Church (antl his other major cx;implc, tlic Synanon drug-treiitinent organization) I'rofessor Akerlof implicitly

212 American Jouriial of Ecotiotuics a?rd Sociolog)] Salience 213

objects to conformist behavior which he regards as irrational (ibid., 10, 11). ~ i e even claims that Synanon “evolved into an organization under the control of a leader who became increasingly insane” (12). I t is evident that Akerlof disap- proves of successful persuasion when it is employed by organizations with goals he considers objectionable.

Although Akerlof might have done a better job at avoiding unnecessarily nor- mative phrasing (at points, the reader gets the distinct impression that Akerlof personally does not Cafe for certain Iirands of conformist), his main point is quite clear. He thinks that many forms o f conformity are incflicient, n o t for society (although this is a possible related issue), but to the conformist himself. Conformity supposedly represents a failure of r;ition;il utility m;iximiz;ition.

Surely Akcrlof would readily agree t1i:it many forms of conl’orriiity ;iIe ellicicnt both to the individual and to the social order. In the most hasic sense, civilization itself depends o n widespread conformity to law and mores. ’]‘here are cerrain segments of the economy where conformity is rewarded. The army and police forces are good examples. An orchestra is a bastion of conformity. ’I’he string instruments must p1;iy on time ;ind i i i iiine lm tliiiigs x c ’ cvcii worst- witli [lie wind players-they are told when and when not to breathe! We assume that Akerlof has no objection to this kind of protluctive conformity.

But this point has 1xo;iclcr imp1ic;itions. Much coiilorniiiy is procluctivc. Con- formity to existing norms of bch:ivior represents the cfforts on t h c part of r:ition:iI individuals to minimize transactions costs, and m:iximize their own utility in the process. Sometimes those conforming inc1ividu;ils m;iy I a t r r come to regret their decision to abide by some particular norm of behavior, but such (expost) regret has no bearing on the rationality of the original decision.

VI I

Who Will Guard the Procrastinators?

ACCORDING TO AKERLOF, tlie “mock1 of cult group 1)eliavior. . . is relevant in understanding politics’ darker side” (1991, 14). Iie proceeds to explain the role of “salience” and “procrastination” in two historical episodes, “Stalin’s Takeover” and the “Vietnam War.”3 Akcrlof conclucles with a discussion of how bureaucracies tend to become popu1;ited with clysfunctional personalities which behave in deviant ways.

But early in his article, Akerlof alludes to the need for government intervention in correcting individual failures to maximize their own utility. In short, individ- uals are dysfunctional, procrastinate and behave irrationally. So the recom- mended policy solution is for dysfunctional, procrastinating, irrational individuals

who happen to hold jobs as government politicians and bureaucrats to step in to restore rationality. This contradiction is hardly worthy of critique.

Akcrlof has collided with an age-old problem philosophers have long wrestled witli. To whit, who will guard the guardians? I f “undue salience” causes indi- viduals to procrastinate and irrationally obey authority why should some people (the guardians) be free of this problem? And if the guardians, along with the rest o f us irrational mortals, suffer from excessively high time preference, how can they ameliorate things for us? For example, assume that Akerlof is correct i n his ;rsscrtion tli;it intlivitluals typicdly S;IVC too little for their old age. I$ut the guarclians, the minions of the state, are also subject to the vagaries of saliency, and also s;ivc too little. ’I‘his circulx rcxsoning is ch:iractc,ristic of hkerlof‘s provoc:rtivc, but Ilawccl, article.

V l l l

Conclusioo

I:ON ~: I~NTI IRI I :S , Kovvriiinciii int(arv(mtioii h;is Iwrn :~~lvoc;itc~l I,y ;I tlivrrsc lot o f intellectuals, including economists, as a panacea. Woefully short-sighted, i l l - ink)rmed indivicluals supposeclly Fail to see the forest of long-run efficiency for tlic trccs o f inimctli;itc giiiii. Public intcrcstccl, long-sighted ;ind well-informed governnicnt officials can intervene to restore the misguicled individual back to the path of wellirre maximization aiicl elliciency.

Akcrlof is the latc*st in ;I long line of such I)roponents, offering a sulxrficially fresh, but ultimately familiar variant on the conventional line of interventionistic reasoning. Unfortunately, Akerlof introduces new jargon (although in this case I)orrowccI f r o m psychology) in to economics with no gain i n theoretical i n s i ~ h t . ~ Contlating later regret with prior irrationality is tantamount to confusing expost with ex ante.

Given the radical uncertainty confronting rational actors in the market, indi- viduals will sometimes make mistakes in their planning. Naturally, such mistakes will often happen after tlie individual receives some misleading information. nut terming this the ellect of “undue salience” ;iclcls nothing to our understanding of how and why that mistake occurred when and where i t did. For this insight, we need to clevelop a true “economics o f error” (following Kirzner).

Notes

1 , hd;iiii Smith ii1:iinI:iincd 111:it tllc “contcinpt o f risk ;lnd thr prcsunipiuous hope of success” caused many 10 overvalue the chance of p i n and under.value the risk of loss (1776/1976, V. 1, 122). Ikirlicr, I lohhcs referred 10 man’s “perverse desire for present profit” (1642, ch. 2, pdragraphs 27 and 32) ;ind suggcstetl h t this led t o an 111idcresliniali011 of h e true value of future returns. Pigou (1920, 25) tliought that ”pcoplc distribute their resources between the present . . . and

214 American Journal of Economics and Sociology Salie t i c e 215

future on the hasis of a wholly irrational preference . . . [the] inevitahle result is that e[forts directed towards the remote future arc starved relatively to those directed towards the . . , present."

2. Unfortunately, there are many other reasonable explanations for whatever level o f poverty of the aged exists which do not presume any lack of self-discipline amongst the young. There are, for instance, many government programs which might he argued to i n i p c t negatively on the elderlyof the 1990s: ruinous monetary mismanagement i n the 192(1s antl 1930s, which greatly enhanced impoverishment o f this age cohort (Friedman and SchwartL, 1963); they have horne an escalating tax burden over their lifetimes, reducing their ahility t o s:tve for the future; antl the impact on the economy of the proliferation of government rex i ih io i is which imp;ictetl them during their most productive years, just t o name a few. 'I'licse devclopnicnts, too, :ire consistent with the poverty r;itcs among tlic rlclcrly nieiilioncd hy hkerlof.

3. Communist I'arty memhers allegedly acquiesced stc1)-l~y-stel) t ( i ht:ilin's i i icrusingly hrutal rule, and gradually committed themselves to "altered standards o f IicIi:ivior" ( I S ) . Siniilarly, President I.ynclon Johnson is clainicd t o Iiave graclu:illy escx1;itecI IIIC* conlli~.t ill Virtn:lni clue t o the role of "salience" related to his brow-heating of dissenters wit l i i i i the Administntion.

4. See Rothhard (1977) for a critique of "welfare m;iximizing constraints oil choice." Jle argues that such a claim is ultimately logically incoherent.

References

Akcrlof, George A,, "l'r(icr~ibtii):ilioii :ind Olwdiciici*," Arwricnrr I;, o i i i i i i i i ( . ~ ~ v I ~ v I ~ ~ , H I , 2, ( IWI ) :

Uanficld, Edward C., "~'rcscnt-Oricntctlness and Crime," AS.SC.LS~JI# / / w <;rimitid. KcMlufiori, Kcfr ihf i ( t r i atid flw /.(,go/ P r o w s , l3:irnett, R:indy li., and I I ; i ~ c ~ l . l o l i n Ill, r d s , , C;inihritlgc, MA: I~al l ingcr, 1977

Becker, Gary 5.. ;ind Murphy. Krvin M.. " A 'I'hcory of 1kition;iI Acklit tion," ,/orrrrrn/ q/I'n/iticnl /k01i011iy96.4 (1988): 675-700

Fisher, Irving, 7%~' 7%rc,orj* rf/ irtcrt~s/ , New York: M:icniill:in, 19.50 Friedman, Milton, and Schwartz, Anna J.. A ilioticlarj* I/islorJf o/lbc /is, /M7- /WX), New York:

Hayek, Friedrich A.. "'l'he Non Sequitur o f the 'I)cpcndcncc liflcct,' " in Sfrrdics i ir I%i/osop/iy,

Hohbes, Thomas, The Cifizeti, London: 1642 lannaccone, Laurence R., "Sacrifice and Stigma: Reducing Free-riding in Cults, Communes, and

Kirzner, Israel. Cottipefilioti aiid Etifrepreiieiirsbip. Chicago: IJ. of Chicago Press, 1973 Milgram, Stanley, Ohcdicttce to AfrthOrftJ? AIZ ,!hpen'mcwlal I k u ; New York: Harper I(r How,

1974 O'Driscoll, Jerry P. , "l'rofessor Ilanfield on ' l ime I lorizen: W h x I las I lc '1':iu~Iit IIs A h o u t Crime?"

in Assessing fhc Crimiiial: Kcsfifiifioti, Kclrihiilioti arid Ihe / . cyyr / I'roccm, Ikirnelt, Randy E., and Ilagel, John I l l , cds.. Camhritlge, MA: Uallingcr, 1977.

Pigou, Arthur, T/ie Economics ofWc'cr/are, 1,ontlon: MacMillan. 1920 Rizzo, Mario J.. "'l'inie Preference, Situ:itional Iktcrniinism. ai ic l Crinic." i n A s w s i t i g fbc Crimiriu/:

K ~ S f i t l i f i ~ J l , Kelrihuliori u t i d fbc l . cp / /~roccsv , lkirnett, llantly I:.. : i n e l l ~ : ~ g c l , ~ o l i n 111, eds., Caml)ridgc, MA: lkill ingcr, 1977

Aothhard, Murray N.. "'l'oward :I Reconstruction of tltil ity and Welfiirc Ikxiiiniiiics," Skin 1:r:incisco: Center for Lihertarian Stirdies, 0ccasion;il k iper #3, 1977

Smith, Adam, AJZ / t iquiry iri lo flit Nrrlrrrc crud CCIIAVCS oft6c W e r i / f h ofMrt iom, Cliic:igo: I J . of Chicago lJ . , 1776/1976

1-19

N;itional Ilurcau of lcononiic l<ese:ircli, 196.5

Politics and I:'cotiomics, New York: Sinioii : i d Se4iubtcr, 1967

Other Collectives,"jotrrtiaf oflblificnl Ecotrottij~ 100.2 (1992). 271-291

Sliglcr, George J,, "'l'he Economics of Informalion,"Jourtio/ of Po/ilical Economy69.3 (1961):

Stigler, George J., and Becker, Gary S., "Dc Gustibus non est Disputandum," American Ecotzomic

Tclser, 1,cster G., "Supply a n d Demand f o r Advertising Messages," American Ecotiomic Keview

213-25

Kcuiew67.1 (1977): 76-90

56.2 (1966): 457-66

Property Taxes in Pennsylvania

1 % ~ A VOTE 01: 26-24, the Pennsylvania Senate this sunimer [ 19941 rejected Ilouse I i i l l 2202, which was in our opinion niislaheled, the "local property vax reform" legislative proposal. The bill would have enabled local governments to levy a 1 percent earned income/net profit tax; or a 1 percent personal income tax; or ;I 1 percent sales and use tax. However, local governments would have had to use 60 percent of those new revenues to reduce property taxes, eliminate nui- saiice taxes, o r 1)otli . Alaritia I i v . t m k of Scotl;iiitl, lb. had souiidcd tlic a1:iriii

last January (see GroundSwell) when a press conference promoting the bill was held by Pennsylvania 1,t. Gov. Mark Singel, Senate Finance Committee Ma- jor i ty Chairman Micliacl L);iwitl:t [l)-Alleglietiy), and Ik!p. Frank 1,;tGrott:t (U- 1,:twrrncc) who is a mrrnhrr of tlic Ilouse hppropri;itions Committee. I iB 2202 had been previously voted ou t o f the Finance Committee.

'l'hc second part o f A ~ ~ I N K I ' S strategy was the introduction of a local option bill cnal>ling thc two-r:ite t;ix t o I x cxtcntletl t o the Pennsylvania boroughs. Prime sponsor of I l l 3 2532 was Rep. Jeffrey Olasz (D-Allegheny), after a two- rate tax was reconimcndcd to hini by a Ilorncstc:tcl consultant. Kep Olasz chairs the Penn. House of Representatives Financial Services Sub-committee of the Business and Economic Development Committee. Among the 27 co-sponsors of I1B 2532 was Rep. Jeffrey Coy (D-Chanibersburg), who is Ilouse Majority Caucus Chairman. Introduction of HB 2532 was preceded by passage by the Chambersburg councilmembers of a resolution requesting that the state pass boroughs en;ibl ing legislation.

GroundSwell